@Override
public void storeSchema(ResourceSchema schema, String location, Job job) throws IOException {
Configuration conf = job.getConfiguration();
DataStorage storage =
new HDataStorage(new Path(location).toUri(), ConfigurationUtil.toProperties(conf));
ElementDescriptor schemaFilePath = storage.asElement(location, schemaFileName);
if (!schemaFilePath.exists() && schema != null) {
try {
new ObjectMapper().writeValue(schemaFilePath.create(), schema);
} catch (JsonGenerationException e) {
log.warn("Unable to write Resource Statistics for " + location);
e.printStackTrace();
} catch (JsonMappingException e) {
log.warn("Unable to write Resource Statistics for " + location);
e.printStackTrace();
}
}
if (printHeaders) {
ElementDescriptor headerFilePath = storage.asElement(location, headerFileName);
if (!headerFilePath.exists()) {
OutputStream os = headerFilePath.create();
try {
String[] names = schema.fieldNames();
String fn;
for (int i = 0; i < names.length; i++) {
fn = ((names[i] == null) ? ("$" + i) : names[i]);
os.write(fn.getBytes("UTF-8"));
if (i < names.length - 1) {
os.write(fieldDel);
} else {
os.write(recordDel);
}
}
} finally {
os.close();
}
}
}
}
/**
* . Given a path, which may represent a glob pattern, a directory, comma separated files/glob
* patterns or a file, this method finds the set of relevant metadata files on the storage system.
* The algorithm for finding the metadata file is as follows:
*
* <p>For each object represented by the path (either directly, or via a glob): If object is a
* directory, and path/metaname exists, use that as the metadata file. Else if parentPath/metaname
* exists, use that as the metadata file.
*
* <p>Resolving conflicts, merging the metadata, etc, is not handled by this method and should be
* taken care of by downstream code.
*
* <p>
*
* @param path Path, as passed in to a LoadFunc (may be a Hadoop glob)
* @param metaname Metadata file designation, such as .pig_schema or .pig_stats
* @param conf configuration object
* @return Set of element descriptors for all metadata files associated with the files on the
* path.
*/
protected Set<ElementDescriptor> findMetaFile(String path, String metaname, Configuration conf)
throws IOException {
Set<ElementDescriptor> metaFileSet = new HashSet<ElementDescriptor>();
String[] locations = LoadFunc.getPathStrings(path);
for (String loc : locations) {
DataStorage storage;
storage = new HDataStorage(new Path(loc).toUri(), ConfigurationUtil.toProperties(conf));
String fullPath = FileLocalizer.fullPath(loc, storage);
if (storage.isContainer(fullPath)) {
ElementDescriptor metaFilePath = storage.asElement(fullPath, metaname);
if (exists(metaFilePath)) {
metaFileSet.add(metaFilePath);
}
} else {
ElementDescriptor[] descriptors = storage.asCollection(loc);
for (ElementDescriptor descriptor : descriptors) {
ContainerDescriptor container = null;
if (descriptor instanceof HFile) {
Path descriptorPath = ((HPath) descriptor).getPath();
Path parent = descriptorPath.getParent();
container = new HDirectory((HDataStorage) storage, parent);
} else { // descriptor instanceof HDirectory
container = (HDirectory) descriptor;
}
// if no custom schema, try the parent directory
ElementDescriptor metaFilePath = storage.asElement(container, metaname);
if (exists(metaFilePath)) {
metaFileSet.add(metaFilePath);
}
}
}
}
return metaFileSet;
}
@Override
public void storeStatistics(ResourceStatistics stats, String location, Job job)
throws IOException {
Configuration conf = job.getConfiguration();
DataStorage storage =
new HDataStorage(new Path(location).toUri(), ConfigurationUtil.toProperties(conf));
ElementDescriptor statFilePath = storage.asElement(location, statFileName);
if (!statFilePath.exists() && stats != null) {
try {
new ObjectMapper().writeValue(statFilePath.create(), stats);
} catch (JsonGenerationException e) {
log.warn("Unable to write Resource Statistics for " + location);
e.printStackTrace();
} catch (JsonMappingException e) {
log.warn("Unable to write Resource Statistics for " + location);
e.printStackTrace();
}
}
}